Joint qualitative and quantitative evaluation of fast image dehazing based on dark channel prior - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Joint qualitative and quantitative evaluation of fast image dehazing based on dark channel prior

Résumé

Haze is an undesirable effect in images caused when atmospheric particles, such as water droplets, ice crystals, dust, or smoke, are lit directly or indirectly by the sun. This effect can be counteracted by image processing, to bring back the details of a hazy image. Unfortunately, the execution time is often long, which prevents deployment in some video or real-time applications. In this paper, we propose to tune several parameters of the Dark Channel Prior method (DCP) algorithm combined with the fast guided filter. We evaluate the optimization in terms of execution time, and quantify the output image quality using different image quality metrics.
Fichier principal
Vignette du fichier
Aksas-SPIE-2022.pdf (22.03 Mo) Télécharger le fichier
Origine : Fichiers éditeurs autorisés sur une archive ouverte

Dates et versions

hal-03815824 , version 1 (09-02-2023)
hal-03815824 , version 2 (17-05-2023)

Identifiants

Citer

Lyes Aksas, Pierre-Jean Lapray, Alban Foulonneau, Laurent Bigue. Joint qualitative and quantitative evaluation of fast image dehazing based on dark channel prior. SPIE PHOTONICS EUROPE, Apr 2022, Strasbourg, France. ⟨10.1117/12.2621566⟩. ⟨hal-03815824v2⟩
49 Consultations
12 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More